[mail-vet-discuss] Draft as of 9/4/2007

[Scott Kitterman]

On Friday 12 October 2007 17:44, Murray S. Kucherawy wrote:
> I incorporated all of these suggestions except:
>
> Eric Allman wrote:
> > Section 5 says that relaying MTAs SHOULD NOT add an A-R header field,
> > even if they actually do check the results and take actions on the
> > basis of the results.  That seems like it can create a mystery.  I
> > suggest either changing the text or adding an explanation for this
> > behavior.
>
> A relaying MTA should only add an A-R header field if the border MTA
> which will receive the message trusts the relaying MTA.  An example I
> added shows how this might be useful.
>
> Thus I guess this should say that a relaying MTA SHOULD NOT add an A-R
> field unless it falls within the trust boundary of the domain to which
> it is relaying.

Isn't trust something the receiver of the relayed mail would decide?  I'm 
thinking of something like a alumni association forwarder wouldn't know if 
receivers chose to trust them or not (and not all receivers would come to the 
same conclusion).

If the relaying MTA has to know anything about how MTAs to which it relays 
feel about it's trustworthiness, I think that's a substantial administrative 
complexity.

I think this should be turned around and left up to receiving MTAs to decide 
if they trust the upstream header or not.

Scott K